Move the shutdown handling to a shutdown function called from
the IPMI core code. That makes for a cleaner shutdown.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Move the shutdown handling to a shutdown function called from
the IPMI core code. That makes for a cleaner shutdown.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
To handle hot remove of interfaces, a lot of rework had to be
done to the locking. Several things were switched over to srcu
and shutdown for users and interfaces was added for cleaner
shutdown.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Counters would not be pegged properly on some errors. Have
deliver_response() return an error so the counters can be
incremented properly.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
The mutex didn't really serve any useful purpose, from what I can
tell, and it would just get in the way. So remove it.
Removing that required a mutex around the default value setting and
getting, so just use the receive mutex for that.
Also pull the fasync stuff outside of the lock for adding the data
to the queue, since it didn't need to be there.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
This is a cleaner interface and the main IPMI panic handler does setup
required by the watchdog handler.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Users of the IPMI code had their own panic handlers, but the
order was not necessarily right, the base IPMI code would
need to handle the panic first, and the user had no way to
know if the IPMI interface could run at panic time.
Add a panic handler to the user interface, it is called if
non-NULL and the interface the user is on is capable of panic
handling. It also cleans up the panic log handling a bit to
reuse the existing interface loop in the main panic handler.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Simplify things by creating one set of message handling data for
setting the watchdog and doing a heartbeat. Rework the locking
to avoid some (probably not very important) races and to avoid
a fairly unlikely infinite recursion.
Get rid of ipmi_ignore_heartbeat, it wasn't used, and use
watchdog_user to tell if we have a working IPMI device below
us.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
If you send a command to another BMC that might take some extra
time, increase the timeouts temporarily.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
By default the retry timeout is 1 second. Allow that to be modified,
primarily for slow operations, like firmware writes.
Also, the timeout was driven by a 1 second timer, so 1 second really
meant between 0 and 1 second. Set the default to 2 seconds so it
means between 1 and 2 seconds.
Also allow the time the interface automatically stays in mainenance
mode to be modified from it's default 30 seconds.
Also consolidate some of the timeout and retry setup.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
more
This patch allows HP PA-RISC servers like rp3410/rp3440 and the HP C8000
workstation with an IPMI controller that predate IPMI 1.5 to use the standard
poweroff or powercycle commands.
These systems firmware don't set the chassis capability bit in the Get
Device ID, but they do implement the standard poweroff and powercycle
commands.
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
This happens when BMC doesn't return any data and the code is trying
to print the value of data[2].
Getting following crash:
[ 484.728410] Unable to handle kernel NULL pointer dereference at virtual address 00000002
[ 484.736496] pgd = ffff0000094a2000
[ 484.739885] [00000002] *pgd=00000047fcffe003, *pud=00000047fcffd003, *pmd=0000000000000000
[ 484.748158] Internal error: Oops: 96000005 [#1] SMP
[...]
[ 485.101451] Call trace:
[...]
[ 485.188473] [<ffff000000a46e68>] msg_done_handler+0x668/0x700 [ipmi_ssif]
[ 485.195249] [<ffff000000a456b8>] ipmi_ssif_thread+0x110/0x128 [ipmi_ssif]
[ 485.202038] [<ffff0000080f1430>] kthread+0x108/0x138
[ 485.206994] [<ffff0000080838e0>] ret_from_fork+0x10/0x30
[ 485.212294] Code: aa1903e1 aa1803e0 b900227f 95fef6a5 (39400aa3)
Adding a check to validate the data len before printing data[2] to fix this issue.
Signed-off-by: Kamlakant Patel <kamlakant.patel@cavium.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Realtek has some sort of "Virtual" IPMI device on the PCI bus as a
KCS controller, but whatever it is, it's not one. Ignore it if seen.
Reported-by: Chris Chiu <chiu@endlessm.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Tested-by: Daniel Drake <drake@endlessm.com>
The IPMI spec states:
The purpose of the SPMI Table is to provide a mechanism that can
be used by the OSPM (an ACPI term for “OS Operating System-directed
configuration and Power Management” essentially meaning an ACPI-aware
OS or OS loader) very early in the boot process, e.g., before the
ability to execute ACPI control methods in the OS is available.
When we are probing IPMI in Linux, ACPI control methods are available,
so we shouldn't be probing using SPMI. It could cause some confusion
during the probing process.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
The IPMI spec states:
The purpose of the SPMI Table is to provide a mechanism that can
be used by the OSPM (an ACPI term for “OS Operating System-directed
configuration and Power Management” essentially meaning an ACPI-aware
OS or OS loader) very early in the boot process, e.g., before the
ability to execute ACPI control methods in the OS is available.
When we are probing IPMI in Linux, ACPI control methods are available,
so we shouldn't be probing using SPMI. It could cause some confusion
during the probing process.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Tested-by: Jiandi An <anjiandi@codeaurora.org>
If platform_device_alloc() then we should return -ENOMEM instead of
returning success.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Use the ARRAY_SIZE macro on a array poweroff_functions to determine
size of the array. Improvement suggested by Coccinelle.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
The cleanup code for an init failure and for a device removal were
quite similar, consolidate all that into one function.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
device_remove_group() was called on any cleanup, even if the
device attrs had not been added yet. That can occur in certain
error scenarios, so add a flag to know if it has been added.
Also make sure we remove the dev if we added it ourselves.
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Cc: stable@vger.kernel.org # 4.15
Cc: Laura Abbott <labbott@redhat.com>
Tested-by: Bill Perkins <wmp@grnwood.net>
Many for coding-style fixes, and update the poll API with the new
type '__poll_t', this is new commit from linux-4.16-rc1.
Signed-off-by: Haiyue Wang <haiyue.wang@linux.intel.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
In preparation to enabling -Wimplicit-fallthrough, mark switch cases
where we are expecting to fall through.
Addresses-Coverity-ID: 1465255 ("Missing break in switch")
Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com>
Cc: Haiyue Wang <haiyue.wang@linux.intel.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
The KCS (Keyboard Controller Style) interface is used to perform in-band
IPMI communication between a server host and its BMC (BaseBoard Management
Controllers).
This driver exposes the KCS interface on ASpeed SOCs (AST2400 and AST2500)
as a character device. Such SOCs are commonly used as BMCs and this driver
implements the BMC side of the KCS interface.
Signed-off-by: Haiyue Wang <haiyue.wang@linux.intel.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Provides a device driver for the KCS (Keyboard Controller Style)
IPMI interface which meets the requirement of the BMC (Baseboard
Management Controllers) side for handling the IPMI request from
host system software.
Signed-off-by: Haiyue Wang <haiyue.wang@linux.intel.com>
[Removed the selectability of IPMI_KCS_BMC, as it doesn't do much
good to have it by itself.]
Signed-off-by: Corey Minyard <cminyard@mvista.com>
This is the mindless scripted replacement of kernel use of POLL*
variables as described by Al, done by this script:
for V in IN OUT PRI ERR RDNORM RDBAND WRNORM WRBAND HUP RDHUP NVAL MSG; do
L=`git grep -l -w POLL$V | grep -v '^t' | grep -v /um/ | grep -v '^sa' | grep -v '/poll.h$'|grep -v '^D'`
for f in $L; do sed -i "-es/^\([^\"]*\)\(\<POLL$V\>\)/\\1E\\2/" $f; done
done
with de-mangling cleanups yet to come.
NOTE! On almost all architectures, the EPOLL* constants have the same
values as the POLL* constants do. But they keyword here is "almost".
For various bad reasons they aren't the same, and epoll() doesn't
actually work quite correctly in some cases due to this on Sparc et al.
The next patch from Al will sort out the final differences, and we
should be all done.
Scripted-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull driver core updates from Greg KH:
"Here is the set of "big" driver core patches for 4.16-rc1.
The majority of the work here is in the firmware subsystem, with
reworks to try to attempt to make the code easier to handle in the
long run, but no functional change. There's also some tree-wide sysfs
attribute fixups with lots of acks from the various subsystem
maintainers, as well as a handful of other normal fixes and changes.
And finally, some license cleanups for the driver core and sysfs code.
All have been in linux-next for a while with no reported issues"
* tag 'driver-core-4.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (48 commits)
device property: Define type of PROPERTY_ENRTY_*() macros
device property: Reuse property_entry_free_data()
device property: Move property_entry_free_data() upper
firmware: Fix up docs referring to FIRMWARE_IN_KERNEL
firmware: Drop FIRMWARE_IN_KERNEL Kconfig option
USB: serial: keyspan: Drop firmware Kconfig options
sysfs: remove DEBUG defines
sysfs: use SPDX identifiers
drivers: base: add coredump driver ops
sysfs: add attribute specification for /sysfs/devices/.../coredump
test_firmware: fix missing unlock on error in config_num_requests_store()
test_firmware: make local symbol test_fw_config static
sysfs: turn WARN() into pr_warn()
firmware: Fix a typo in fallback-mechanisms.rst
treewide: Use DEVICE_ATTR_WO
treewide: Use DEVICE_ATTR_RO
treewide: Use DEVICE_ATTR_RW
sysfs.h: Use octal permissions
component: add debugfs support
bus: simple-pm-bus: convert bool SIMPLE_PM_BUS to tristate
...
Pull IPMI updates from Corey Minyard:
"Small fixes for various things, been sitting in next for a while (some
a long time)"
* tag 'for-linus-4.16-1' of git://github.com/cminyard/linux-ipmi:
ipmi_ssif: Remove duplicate NULL check
ipmi/powernv: Fix error return code in ipmi_powernv_probe()
ipmi: use dynamic memory for DMI driver override
ipmi/ipmi_powernv: remove outdated todo in powernv IPMI driver
ipmi: Clear smi_info->thread to prevent use-after-free during module unload
ipmi: use correct string length
ipmi_si: Fix error handling of platform device
ipmi watchdog: fix typo in parameter description
ipmi_si_platform: Fix typo in parameter description
Pull poll annotations from Al Viro:
"This introduces a __bitwise type for POLL### bitmap, and propagates
the annotations through the tree. Most of that stuff is as simple as
'make ->poll() instances return __poll_t and do the same to local
variables used to hold the future return value'.
Some of the obvious brainos found in process are fixed (e.g. POLLIN
misspelled as POLL_IN). At that point the amount of sparse warnings is
low and most of them are for genuine bugs - e.g. ->poll() instance
deciding to return -EINVAL instead of a bitmap. I hadn't touched those
in this series - it's large enough as it is.
Another problem it has caught was eventpoll() ABI mess; select.c and
eventpoll.c assumed that corresponding POLL### and EPOLL### were
equal. That's true for some, but not all of them - EPOLL### are
arch-independent, but POLL### are not.
The last commit in this series separates userland POLL### values from
the (now arch-independent) kernel-side ones, converting between them
in the few places where they are copied to/from userland. AFAICS, this
is the least disruptive fix preserving poll(2) ABI and making epoll()
work on all architectures.
As it is, it's simply broken on sparc - try to give it EPOLLWRNORM and
it will trigger only on what would've triggered EPOLLWRBAND on other
architectures. EPOLLWRBAND and EPOLLRDHUP, OTOH, are never triggered
at all on sparc. With this patch they should work consistently on all
architectures"
* 'misc.poll' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (37 commits)
make kernel-side POLL... arch-independent
eventpoll: no need to mask the result of epi_item_poll() again
eventpoll: constify struct epoll_event pointers
debugging printk in sg_poll() uses %x to print POLL... bitmap
annotate poll(2) guts
9p: untangle ->poll() mess
->si_band gets POLL... bitmap stored into a user-visible long field
ring_buffer_poll_wait() return value used as return value of ->poll()
the rest of drivers/*: annotate ->poll() instances
media: annotate ->poll() instances
fs: annotate ->poll() instances
ipc, kernel, mm: annotate ->poll() instances
net: annotate ->poll() instances
apparmor: annotate ->poll() instances
tomoyo: annotate ->poll() instances
sound: annotate ->poll() instances
acpi: annotate ->poll() instances
crypto: annotate ->poll() instances
block: annotate ->poll() instances
x86: annotate ->poll() instances
...
Fix to return a negative error code from the request_irq() error
handling case instead of 0, as done elsewhere in this function.
Fixes: dce143c338 ("ipmi/powernv: Convert to irq event interface")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Since the IPMI core now queries device IDs dynamically, we no longer
need this todo for implementing this in the powernv driver.
Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
During code inspection, I found an use-after-free possibility during unloading
ipmi_si in the polling mode.
If start_new_msg() is called after kthread_stop(), the function will try to
wake up non-existing kthread using the dangling pointer.
Possible scenario is when a new internal message is generated after
ipmi_unregister_smi()[*1] and remains after stop_timer_and_thread()
in clenaup_one_si() [*2].
Use-after-free could occur as follows depending on BMC replies.
cleanup_one_si
=> ipmi_unregister_smi
[*1]
=> stop_timer_and_thread
=> kthread_stop(smi_info->thread)
[*2]
=> poll
=> smi_event_handler
=> start_new_msg
=> if (smi_info->thread)
wake_up_process(smi_info->thread) <== use-after-free!!
Although currently it seems no such message is generated in the polling mode,
some changes might introduce that in thefuture. For example in the interrupt
mode, disable_si_irq() does that at [*2].
So let's prevent such a critical issue possibility now.
Signed-off-by: Yamazaki Masamitsu <m-yamazaki@ah.jp.nec.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
gcc-8 reports
drivers/char/ipmi/ipmi_msghandler.c: In function
'panic_op_write_handler':
./include/linux/string.h:245:9: warning: '__builtin_strncpy' specified
bound 16 equals destination size [-Wstringop-truncation]
drivers/char/ipmi/ipmi_watchdog.c: In function 'set_param_str':
./include/linux/string.h:245:9: warning: '__builtin_strncpy' specified
bound 16 equals destination size [-Wstringop-truncation]
We need one less byte or call strlcpy() to make it a nul-terminated
string.
Signed-off-by: Xiongfeng Wang <xiongfeng.wang@linaro.org>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Cleanup of platform devices created by the IPMI driver was not
being done correctly and could result in a memory leak. So
create a local boolean to know how to clean up those platform
devices.
Reported-by: David Binderman <dcb314@hotmail.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Fix typo in parameter description.
Fixes: 95e300c052 ("ipmi: Make the DMI probe into a generic platform probe")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Pull IPMI fixes from Corey Minyard.
* tag 'for-linus-4.15-2' of git://github.com/cminyard/linux-ipmi:
ipmi_si: fix crash on parisc
ipmi_si: Fix oops with PCI devices
ipmi: Stop timers before cleaning up the module
This patch fixes ipmi crash on parisc introduced in the kernel 4.15-rc.
The pointer io.io_setup is not initialized and thus it causes crash in
try_smi_init when attempting to call new_smi->io.io_setup.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
